Dynamic composite multi-gradient systems for advanced MRI Abstract: This project will use a novel composite gradient system to test that images of the fine structure of inner and middle ear anatomy obtained with magnetic resonance imaging (MRI) and, ultimately, the diagnosis of inner and middle ear pathology can be greatly improved by gradient performance that is increased over the conventional typical amplitudes of 40 to 45 mT/m and slew rates of 200T/m/s. The gradient system will combine the system body gradients simultaneously with an insert head gradient to achieve gradient performance that is nearly a factor of 3 greater in speed and amplitude than the body gradients alone. Because pathologies of the inner ear require very high spatial resolution that is often beyond the capabilities of conventional MRI scanners, the experiments proposed in this study to detect these pathologies are ideal tests for the improved performance of composite gradients. By determining the cause of sudden sensorineural hearing loss the appropriate treatment could be specified early in the disease management. An objective test for endolymphatic hydrops would allow improved diagnosis and monitoring of therapy in Meniere's and related disorders. By accurately discriminating between residual cholesteatoma and other non-tumor tissues of the middle ear, this project could eliminate many unnecessary surgeries. To the best of our knowledge, this is the first work to systematically evaluate the value of very high gradient performance in diagnostic studies of the middle and inner ear. In this study, the body and local gradients will be used simultaneously with equivalent waveform shapes and timing. This work will advance the technology available for clinical MRI by demonstrating: 1) the flexibility and potential performance improvement of composite gradient systems over conventional single gradient systems and 2) that no image quality is lost when using only the body gradients with the insert gradients in position. Thus, the insert gradients can be kept in position for clinical whole brain studies and be available for use on sequences that require increased gradient performance. These observations have never before been demonstrated.